The Bruised Brother elsewhere on the Internet

I'm really proud of this comment I made on this post:

“advanced” stats are called “advanced” for reasons that i couldn’t tell you. we all seem comfortable talking about goalie save percentages and goals against average, yet we bristle at the thought of what a player’s points per 60 or goals for % is.

as J.J. said and i would like to emphasize, advanced stats are absolutely NOT crucial to appreciating and enjoying the game. i had a lot of fun watching hockey before i ever heard the word Corsi, but considering the direction that hockey analytics seems very likely headed, i decided to familiarize myself with them. you likely won’t ever see me searching out Corsi or any of those numbers myself because that’s just not how i describe the game or interact with other fans. i’ll inevitably say something dumb in the future, but generally, i just listen when the discussions move toward player comparisons using these metrics. considering i’m also learning to play hockey myself, i’m also slowly gaining a more complete understanding of tactics. granted, that process will be long, slow, and arduous, but that’s just how i engage with the game.

as for the stats themselves, i do find it helpful to distinguish between the different kind of non-traditional stats. most of the traditional stats are just simply counting stats—goals, assists, points, plus-minus, PIM, and the like. goalie stats can help ease you into how certain metrics are calculated: GAA is simply GA/60 minutes played for a goalie, or how many goals a goalie gives up on average per game. when talking about rate stats for skaters, these usually manifest as G/60 or P/60 or A1/60 for primary assists and A2/60 for secondary assists. you’ve got percentages, which there are a ton of: shooting %, save %, goals for/ against %. then you have the possession stats Corsi and Fenwick and all the various metrics associated with them like quality of competition (QUALCOMP). this is a lot of name-dropping, i admit, so don’t fret if you didn’t understand it the first time. the thing i hope you take away from this paragraph is that “advanced” stats encompass a lot of ways of looking at the numbers you’re already familiar with, like goals and points, and seeing what patterns emerge as well as looking at other numbers like shots and shot differentials. they’re just different ways of looking at players and the game.

the inevitable strawman that seems to pop up in the debate over advanced stats concerns their predictive power. when someone says that Corsi is the best and most reliable indicator of long-term predictability or repeatability for players and teams, they mean it exactly as it sounds: it’s the best long-term indicator. compiling all the players’ Corsi and goalie stats and throwing them into a machine can give Vegas betting odds, but it won’t tell us if the Red Wings are going to win opening night against Buffalo, or the Winter Classic against the Maple Leafs, or Lidstrom night against the Avalanche. dealing with hockey on a game by game and period by period basis, the numbers have almost no value because periods and games are examples of small sample sizes that swing the numbers into a wild variance. the numbers can tell us who will likely make the playoffs, and even to a certain extent how far teams will go in the playoffs, but even in the long-term, there is still a level of uncertainty in their predictive power over the influence of the outcome of a hockey game, season, or playoff series.

so to answer your questions: advanced stats usually refer to anything that doesn’t show up on NHL.com’s stats pages that can still tell us something useful about players and teams. they’re important for the information they tell us, but always keep in mind that no one number can possibly capture everything in a hockey game. even advanced stats need context and proper interpretation to be utilized to their full analytical potential. but they’re not absolutely crucial to watching and enjoying a hockey game or season.

--uvgt2bkdnme